• Status: Solved
  • Priority: Medium
  • Security: Public
  • Views: 289
  • Last Modified:

Why a single update, delete and insert operations are atomic, even without using Begin Transaction ?


I have a doubt about the internal behavior of how SQL Server handle errors in update/delete/insert operations.

If say, I have an "Update" statement, that will update 10000 records and in the middle of the update, some error happens, like a check constraint violation, I review the table and non of the records where updated. That's very nice, right ? Even without adding a Begin Transaction / Update|Delete|Insert ... / Commit transaction, the operation is always atomic.
So the question is how does it do it ?

I have some guess. First, could it be starting an internal/hidden transaction  ? If so, is that transaction the same kind of transaction we use in our code or do they have their own special internal transaction to handle this ?
Or, if this is not a transaction at all, then what is it ?

Thanks in advance,

  • 2
1 Solution
Your guess is correct, and it's called an "implicit" transaction. It's the same kind of transaction as when you write BEGIN TRAN in T-SQL.
fischermxAuthor Commented:

Thanks for your answer,  jdlambert1.

However, let me clarify this for other readers. The term "Implicit Transaction" has slightly different meaning. It is when you never have to call "begin transaction", you just say commit or rollback. You can do some insertions, then some udpates and so on, and then, you just call commit and all of them are commited.
Each next transaction begins exactly where the prior ended, without having to begin it explicitly, that's why they're called implicit :)

I understand that what I'm seeing is a sort-of implicit transaction, since it looks like a transaction and behaves like a transaction and since I didn't started, then it is "implicit" as you say. But I wanted clarify the term.

So, my question was what happens inside a SINGLE update statement that handles hundreds of records and it fails in the middle of it. I guess I may find a deeper anwser in Kalen Delaney's SQL Server book, which is on the way from amazon, but I wanted to gather some opinions here in the meantime.


It's the same meaning to my way of thinking: a transaction must be explicit or implicit, and if your code doesn't state the beginning of the transaction, that's implicit. All the major RDBMS's use transactions for all single statements by design, for the sake of data integrity, obviously. If you want independent behavior, you have to put each record (or group of records you want done as a unit) in separate statements.

Inside SQL Server is a great book, and I hope they get Kalen to update it again for 2005.
Question has a verified solution.

Are you are experiencing a similar issue? Get a personalized answer when you ask a related question.

Have a better answer? Share it in a comment.

Join & Write a Comment

Featured Post

Free Tool: ZipGrep

ZipGrep is a utility that can list and search zip (.war, .ear, .jar, etc) archives for text patterns, without the need to extract the archive's contents.

One of a set of tools we're offering as a way to say thank you for being a part of the community.

  • 2
Tackle projects and never again get stuck behind a technical roadblock.
Join Now